AI Transparency Deadline Approaching

On August 2, several provisions of the EU AI Act become enforceable, focusing on transparency for AI systems interacting with people and the machine-readable watermarking of AI-generated content. Deployers of deepfake-generating AI systems must disclose such content, with exceptions for legal investigations and artistic works. Some AI Act obligations, particularly for high-risk systems, have been postponed following the approval of the Digital Omnibus on AI.

• Transparency is required for AI systems that interact directly with people, unless obvious or used for law enforcement.
• AI-generated content (audio, image, video, text) must be marked in a machine-readable format, with exceptions for standard editing, assistive functions, or law enforcement.
• A grace period until December 2 applies to AI systems already on the market before August 2 for content generation watermarking.
• Deployers of emotion recognition or biometric categorization systems must inform individuals and comply with data protection regulations, with exceptions for law enforcement.
• Disclosure is required for deepfake content, unless authorized for criminal investigations or part of artistic/satirical works.
• Some AI Act obligations, including rules for high-risk AI systems, have been postponed.
  https://bender.layer3.press/articles/65eed0ca-7b06-4c4d-8bbb-23e73676a730